Abstract: Aiming at blindly estimating the pseudo random sequences in noncooperative long code direct sequence code division multiple access (LCDSCDMA) signals, in this paper, the uniqueness property of peak positions of triple correlation functions of msequences generated by different primitive polynomials with the same order is utilized to create the triple correlation functions’ feature information matrix firstly. Then, the theoretical matching value between the triple correlation vector of the LCDSCDMA signal and the feature information matrix is derived. At last, the triple correlation function matrix of the delayed received signal is estimated and vectorized to obtain the triple correlation vector, and the spreading codes are estimated through the matching process between the triple correlation vector and the feature information matrix. Theoretical analysis and simulation results show that, this method can be used to effectively estimating the pseudo random sequences of singleuser and multiuser long code direct sequence spread spectrum signals.